US7702002B2  Rapid acquisition methods and apparatus for GPS signals  Google Patents
Rapid acquisition methods and apparatus for GPS signals Download PDFInfo
 Publication number
 US7702002B2 US7702002B2 US10936177 US93617704A US7702002B2 US 7702002 B2 US7702002 B2 US 7702002B2 US 10936177 US10936177 US 10936177 US 93617704 A US93617704 A US 93617704A US 7702002 B2 US7702002 B2 US 7702002B2
 Authority
 US
 Grant status
 Grant
 Patent type
 Prior art keywords
 data
 set
 frequency
 signal
 frequency samples
 Prior art date
 Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
 Expired  Fee Related, expires
Links
Classifications

 G—PHYSICS
 G01—MEASURING; TESTING
 G01S—RADIO DIRECTIONFINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCEDETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
 G01S19/00—Satellite radio beacon positioning systems; Determining position, velocity or attitude using signals transmitted by such systems
 G01S19/01—Satellite radio beacon positioning systems transmitting timestamped messages, e.g. GPS [Global Positioning System], GLONASS [Global Orbiting Navigation Satellite System] or GALILEO
 G01S19/13—Receivers
 G01S19/24—Acquisition or tracking or demodulation of signals transmitted by the system
 G01S19/29—Acquisition or tracking or demodulation of signals transmitted by the system carrier, including Doppler, related

 G—PHYSICS
 G01—MEASURING; TESTING
 G01S—RADIO DIRECTIONFINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCEDETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
 G01S19/00—Satellite radio beacon positioning systems; Determining position, velocity or attitude using signals transmitted by such systems
 G01S19/01—Satellite radio beacon positioning systems transmitting timestamped messages, e.g. GPS [Global Positioning System], GLONASS [Global Orbiting Navigation Satellite System] or GALILEO
 G01S19/13—Receivers
 G01S19/24—Acquisition or tracking or demodulation of signals transmitted by the system
 G01S19/30—Acquisition or tracking or demodulation of signals transmitted by the system code related

 H—ELECTRICITY
 H04—ELECTRIC COMMUNICATION TECHNIQUE
 H04B—TRANSMISSION
 H04B1/00—Details of transmission systems, not covered by a single one of groups H04B3/00  H04B13/00; Details of transmission systems not characterised by the medium used for transmission
 H04B1/69—Spread spectrum techniques
 H04B1/707—Spread spectrum techniques using direct sequence modulation

 G—PHYSICS
 G01—MEASURING; TESTING
 G01S—RADIO DIRECTIONFINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCEDETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
 G01S19/00—Satellite radio beacon positioning systems; Determining position, velocity or attitude using signals transmitted by such systems
 G01S19/01—Satellite radio beacon positioning systems transmitting timestamped messages, e.g. GPS [Global Positioning System], GLONASS [Global Orbiting Navigation Satellite System] or GALILEO
 G01S19/03—Cooperating elements; Interaction or communication between different cooperating elements or between cooperating elements and receivers
 G01S19/09—Cooperating elements; Interaction or communication between different cooperating elements or between cooperating elements and receivers providing processing capability normally carried out by the receiver

 G—PHYSICS
 G01—MEASURING; TESTING
 G01S—RADIO DIRECTIONFINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCEDETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
 G01S19/00—Satellite radio beacon positioning systems; Determining position, velocity or attitude using signals transmitted by such systems
 G01S19/01—Satellite radio beacon positioning systems transmitting timestamped messages, e.g. GPS [Global Positioning System], GLONASS [Global Orbiting Navigation Satellite System] or GALILEO
 G01S19/13—Receivers
 G01S19/24—Acquisition or tracking or demodulation of signals transmitted by the system
 G01S19/25—Acquisition or tracking or demodulation of signals transmitted by the system involving aiding data received from a cooperating element, e.g. assisted GPS

 H—ELECTRICITY
 H04—ELECTRIC COMMUNICATION TECHNIQUE
 H04B—TRANSMISSION
 H04B2201/00—Indexing scheme relating to details of transmission systems not covered by a single group of H04B3/00  H04B13/00
 H04B2201/69—Orthogonal indexing scheme relating to spread spectrum techniques in general
 H04B2201/707—Orthogonal indexing scheme relating to spread spectrum techniques in general relating to direct sequence modulation
 H04B2201/70715—Orthogonal indexing scheme relating to spread spectrum techniques in general relating to direct sequence modulation with applicationspecific features
Abstract
Description
This application claims priority from U.S. provisional patent application Ser. No. 60/540,120, filed Jan. 28, 2004.
1. Field
This application relates to apparatus and methods for computing the position of a mobile device by use of wireless signals, such as from GPS systems.
2. Related Art
Position location devices are becoming increasingly popular, not only for ships at sea and adventurers in the wilderness, but also for the average person with a cell phone. The increasing popularity of personal position location devices has encouraged development of rapid, high sensitivity methods for acquiring the signals used to determine position.
Position location technologies typically utilize wireless signals concurrently transmitted from known locations to determine position. In GPS systems, these signals are concurrently transmitted from a multiplicity of satellites at a known time, and with a predefined frequency. On the ground, a GPS receiver acquires a signal from each satellite within its view of the sky. The times of arrival of the signals along with the exact location of the inview satellites and the exact times the signals were transmitted from each satellite are used to locate the position of the GPS receiver via a trilateration calculation.
The acquisition of signals from the GPS satellites can be difficult due to a number of factors. For example, GPS signals are transmitted at a relatively low power, and from a great distance. By the time the GPS signals travel from earth orbit to a receiver, their initially low power has been greatly reduced, rendering the signal extremely weak at the receiver. The received signal levels may be further weakened by building blockage effects, such as may occur during indoor reception or reception in urban canyon environments.
There are two principal functions of GPS receiving systems: (1) computation of the pseudoranges to the various GPS satellites, and (2) computation of the position of the receiving platform using these pseudoranges and satellite timing and ephemeris (position) data. The pseudoranges measure the time delays (or equivalently the ranges) between the satellites and the GPS receiver with a bias due to the local clock. In conventional autonomous GPS receivers, the satellite ephemeris and time of transmission data are extracted from the GPS signal once it is acquired and tracked. Collecting this information normally takes a relatively long time (30 seconds to several minutes) and must be accomplished with a good received signal level in order to achieve low error rates.
Virtually all known GPS receivers utilize correlation methods, or their mathematical equivalents, to compute pseudoranges. These correlation methods are performed in real time, often with hardware correlators. GPS signals contain high rate repetitive signals that are modulated in accordance with special sequences or “codes”, called pseudorandom (PN) sequences. The codes available for civilian applications are called C/A codes, and are utilized to provide a binary phasereversal rate, or “chipping” rate, of 1.023 MHz and a repetition period of 1023 chips for a code period of 1 msec. The pseudorandom sequences in the GPS system belong to a family known as “Gold codes”. Each GPS satellite broadcasts a signal with a unique Gold code.
For brevity, in the following discussion, we may use the terminology that a signal “contains a pseudorandom sequence” (or code), by which we mean it contains a waveform that is modulated in accordance with a pseudorandom sequence, or code The length of a frame of a pseudorandom sequence is the number of symbols of the sequence before it repeats. By the duration (in time) of a pseudorandom sequence we mean the duration of the waveform modulated in accordance with the pseudorandom sequence. Similarly, when we say frame rate of a pseudorandom sequence we mean the repetition rate of a waveform modulated in accordance with the pseudorandom sequence. It should be clear from the context whether the term pseudorandom sequence refers to a sequence of numbers or a waveform modulated according to such a sequence of numbers.
For a signal received from a given GPS satellite, following a downconversion process to baseband, the signal is then correlated with a reference signal. For example, a simple correlation receiver multiplies the received signal by a locally generated reference signal containing a stored replica of the appropriate Gold code contained within its local memory, and then integrates, (e.g. lowpass filters) the product in order to obtain an indication of the presence of the signal. This process is termed a coherent processing operation. Often the results of multiple successive coherent processing operations are detected (e.g., their magnitudes are determined) and combined in order to provide a final correlation output with higher fidelity. Hence, the initial correlation operations are coherent processing ones, and the latter combining step is called “noncoherent” or “incoherent” processing. In summary, correlation may include only coherent processing, or it may also include incoherent processing of the correlation output of two or more coherent processing steps. The final output of a succession of such correlation processing steps is termed a “final correlation output”.
In this discussion a simple individual correlation process results in a single number (perhaps complex). However, in many cases of interest a multiplicity of such numbers are computed corresponding to different reference sequences (e.g. delayed versions), either serially or in parallel, by performing similar operations. We refer to such a set of numbers as a “correlation series”. The final result of combining one or more successive correlation series is referred to as a “final correlation series.”
By sequentially adjusting the relative timing of this stored replica relative to the received signal, and observing when high energy occurs in the resulting final correlation series, a simple receiver can determine the time delay between the received signal and a local clock. This time delay, modulo the 1 millisecond code period, is termed the “code phase.” The initial determination of the presence of such an output is termed “acquisition.” Once acquisition occurs, the process enters the “tracking” phase in which the timing of the local reference is adjusted in small amounts in order to maintain a high correlation output. The correlation output during the tracking phase may be viewed as the GPS signal energy with the pseudorandom code removed, or, in common terminology, “despread.” This signal is narrow band, with bandwidth commensurate with a 50 bit per second (bps) binaryphaseshiftkeyed data signal that is superimposed on the GPS waveform.
Unfortunately, the correlation acquisition process is very time consuming, especially if received signals are weak. To improve acquisition time, most conventional GPS receivers utilize a multiplicity of correlators (up to 12 typically) which allows a parallel search for correlation peaks.
Some prior GPS receivers have used FFT techniques to determine the Doppler frequency of the received GPS signal. These receivers utilize conventional correlation operations to despread the GPS signal and provide a narrow band signal with bandwidth typically in the range of 10 kHz to 30 kHz. The resulting narrow band signal is then Fourier analyzed using FFT algorithms to determine the carrier frequency. The determination of such a carrier simultaneously provides an indication that the local PN reference is adjusted to the correct code phase of the received signal and provides an accurate measurement of carrier frequency. This frequency may then be utilized in the tracking operation of the receivers.
As mentioned above, the increasing popularity of personal position location devices has encouraged development of rapid, high sensitivity methods for acquiring the GPS signals used to determine position. One method, for example, uses an FFT algorithm to compute pseudoranges at a central processing location rather than at a mobile unit. According to that method, a snapshot of data is collected by a GPS receiver and then transmitted over a data link to a remote receiver where it undergoes FFT processing in order to compute the final correlation series. However, typically, only a single forward and inverse Fast Fourier Transform (corresponding to four PN periods) are computed to perform the set of correlations.
Another approach uses fast Fourier transform methods for acquiring GPS signals includes digitizing, storing, and processing a long block of raw data. For example, data corresponding to a onesecond interval can be digitized and then processed locally using an FFTbased signal processing method to acquire the GPS signals present within this captured data block. In this method a multiplicity of FFT operations are performed, each producing a correlation series, and the results undergo both coherent and incoherent processing operations in order to produce the final correlation series.
Unfortunately, the GPS signal acquisition approach in such systems becomes less efficient when performing very long coherent integrations, such as those exceeding a period of one data bit (e.g., 20 GPS frames, which equals 20 milliseconds of time); the efficiency loss is especially great when the GPS carrier frequency uncertainty is large. Furthermore, it may be noted that in current GPS receiving systems coherent integration over periods exceeding one data bit requires that the GPS receiver have a priori knowledge of the bit sequence. Therefore, coherent integration over periods exceeding one data bit is normally done by transmitting such information from a server to a mobile station. This general approach has been standardized in several cellular communications standards, including IS95, CDMA2000, GSM, and UMTS standards. It should be noted that future GPS systems may have modes or signaling components that transmit PN signals containing no data. Hence in the future, there may be situations in which long coherent integration may be accomplished without the need for the receiver to have a priori knowledge of such a bit sequence.
Other prior approaches to coherent processing may be useful when (1) long coherent integration is required, (2) a search over a wide Doppler range is required, and (3) a code phase search must be performed over the full 1023 chips of each GPS signal to be processed. Such prior approaches, however, have a number of limitations and restrictions. For example, these algorithms may require processing data as a twodimensional array and also limit the extent over which a Doppler search may be efficiently performed.
A method and apparatus is described herein for receiving and processing one or more signals transmitted from a plurality of transmitters at predetermined frequencies. Each of the transmitted signals includes a waveform coded according to a periodicallyrepeating sequence that uniquely identifies the transmitter that sent each respective signal. The received signals are used in determining a location of the receiver. In one embodiment the transmitters comprise a plurality of GPS satellites that transmit GPS signals at a GPS frequency, each GPS satellite transmitting a waveform coded according to a unique periodicallyrepeating sequence. The signal's code phase offset at the receiver is found, and using this information from a number of transmitters, the receiver's position may be fixed using GPS algorithms.
As will be evident from the following description, higher sensitivity and higher processing speed can be achieved by performing FFT operations on the observed data; in conjunction with the FFTs, special pruning operations are used based upon a hypothesized residual frequency error to significantly reduce the total number of calculations and therefore significantly reduce processing time.
Particularly, at the receiver, the signal at the predetermined frequency is observed and digitized over a predefined period of time corresponding to at least two repetitions of the periodicallyrepeating sequences (two frames). One of the plurality of transmitters is hypothesized, and a set of reference frequency samples corresponding to the hypothesized transmitter is provided. A first subset of the digitized data is selected of duration at least equal to two frames, thus defining a block of data. A set of data frequency samples is then calculated from this block, such as by using Fourier transform techniques. A first residual frequency is hypothesized, and then the data frequency samples are pruned responsive to the hypothesized first residual frequency to provide a periodicallyspaced first subset of the data frequency samples. The first subset of the data frequency samples and the reference frequency samples are further processed (typically with a multiplication and inverse FFT procedure) to provide a first correlation data series. This procedure may then be repeated upon additional data blocks (typically contiguous) and the multiple correlation series so found may be detected and added together to form a final correlation series. This latter series is then searched to identify a signal match, typically by looking for a strong peak in the final correlation series. If a matched signal is found, a code phase offset is determined from the final correlation data series; however if a matched signal is not found, then another residual frequency may be hypothesized, and the process repeated, typically using the same sets of data frequency samples and reference frequency samples, to search for a signal match. Similar processing proceeds until a signal match is found or until enough residual frequencies have been hypothesized without finding a match to assume that the signal from the hypothesized transmitter cannot be acquired.
Typically, there may be a number of transmitters that may be viewable by the receiver, and this process may be repeated for each of such transmitters in order to identify the signals and determine a code phase offset from each transmitter, if possible.
Many different embodiments can be implemented. In one embodiment, the pruning step further comprises selecting a subset of the data frequency samples, the subset including a plurality of samples having indices spaced relative to one another by an integer K, wherein K is the number of frames of the PN sequence in the data block.
In one embodiment, the method further comprises multiplying the subset of the data frequency samples with a set of reference frequency samples, to form a set of weighted frequency samples.
The reference frequency samples are obtained by any suitable method, for example the receiver may perform a discrete Fourier Transform (DFT) operation upon one or more periods of the periodicallyrepeated sequences to define the reference frequency samples, the reference frequency samples may be precalculated for each transmitter and stored in the receiver, or the reference frequency samples may be downloaded from a server, such as the PDE described herein.
The step of performing a correlation operation may include performing an inverse DFT upon the set of weighted frequency samples, in order to produce a correlation data series.
Each data block may have a size that corresponds to an integral number of repetitions of the periodicallyrepeating sequences two or greater, for example, five, ten, twenty or more. In some embodiments, the data block has a size within a range of about five to twenty repetitions of the periodicallyrepeating sequences. In other embodiments the data block has a size on the order of one hundred such repetitions.
The above method may be implemented in suitable hardware and/or software in the receiver, and/or on one or more servers in the wireless network. For example, some functions may be implemented in the receiver, and some functions may be implemented in a position determination entity (PDE).
The apparatus and method disclosed herein is particularly useful for assisted GPS (“AGPS”) systems in which the communication system, providing assistance information to the GPS receiver, is unsynchronized, as is the case for GSM and UMTS cellular standards. Although in synchronized communication systems such as CDMA2000 standard, the requirements imposed upon the code phase search are greatly eased, there would still be benefits from using the improved algorithms discussed herein.
Reference is now made to the accompanying drawings, wherein:
In the various figures of the drawing, like reference numerals denote like or similar parts.
Glossary of Terms and Acronyms
The following terms and acronyms are used throughout the detailed description:
AGPS: Assisted GPS. A location technology in which special assistance to the GPS acquisition process is provided by a location server, which can reduce acquisition time and improve sensitivity.
Base Station or BTS: Base Transceiver Station. A fixed station used for communicating with mobile stations. Includes antennas for transmitting and receiving wireless communication signals.
CDMA: Code Division Multiple Access. A highcapacity digital wireless technology that was pioneered and commercially developed by QUALCOMM™ Incorporated.
Chip: A symbol, also the smallest part of a modulation (e.g., a phase reversal) of a pseudorandom (PN) sequence.
Chip rate: Symbol transmission rate.
Code phase: The relative timing between a frame boundary of a received signal modulated by a PN sequence contained within a received GPS signal and the frame boundary of a locallygenerated signal modulated by PN reference sequence. For the U.S. GPS system, the code phase is normally a number that is between zero and the duration of one PN frame, i.e., one millisecond in the civilian GPS signal. Code phase may be considered the timing information extracted by the receiver that is used to determine position.
Coherent Processing (or Coherent Integration): The technique of processing a data block with linear processing, that is, without the use of nonlinear (e.g., detection) operations.
Correlation: The process of comparing a received signal and a reference signal. Correlation operations may include coherent integration and incoherent integration. In its simplest form a correlation operation results in a single number indicating the strength of the comparison. The output of a multiplicity of correlation operations as a function of different reference signals (e.g. delayed versions of one another) is termed a “correlation series”. Multiple correlation series may be combined either coherently or incoherently (following nonlinear detection operations). The final result computed after completing all coherent and incoherent processing is called a “final correlation series.”
Detection: The process of operating upon a set of samples by use of a nonlinear operation, typically a magnitude or a magnitudesquared operation, to determine the power in each data sample. For example, if each data sample is represented in an inphase and quadrature (IQ) component format, or in a “complex” format in which the inphase component is associated with the real component of a data word and the quadrature component is associated with the imaginary component, then the detection process may calculate the magnitude or magnitudesquared. In many cases, the magnitude operations are normal operations associated with complex numbers.
DFT: Discrete Fourier Transform.
FFT: Fast Fourier Transform. An efficient technique for calculating the DFT of finiteduration sequences. Note that the “frequency” data samples constructed from a DFT or an FFT algorithm are identical, and we may use the terminology “DFT data” and “FFT data” interchangeably. In most situations an FFT algorithm is used, however, to compute the DFT.
GPS: Global Positioning System. A technique utilizing distance measurements to GPS satellites to determine threedimensional location (including altitude). Although the term GPS is often used to refer to the U.S. Global Positioning System, the meaning of this term includes other global positioning systems, such as the Russian Glonass System and the planned European Galileo System. In some systems, the term Satellite Positioning System (SPS) is used in substitution for the GPS terminology. For illustrative purposes, the description herein is in terms of the current U.S. GPS System. It should be obvious, however, that the principles herein described may be applied to a variety of SPS systems that utilize similar signaling formats, as well as to future variations of the U.S. GPS System.
GPS SA message: The GPS Sensitivity Assistance message of the IS801 protocol includes predicted bits of the currently visible SV's. The bits are predicted by the PDE and sent from the PDE to the MS in OTA (“overtheair”) format. These data bits are sometimes called (predicted) navigational message bits.
GPS fix: The end result of a process of measurements and subsequent computations by which the location of the GPS user is determined.
GSM: Global System for Mobile, another widelyused digital wireless technology.
Incoherent Processing: Combining several adjacent data sets, or series, by performing a nonlinear detection operation, followed by a summing operation, for reasons such as to improve the signaltonoise ratio. For example, incoherent processing may include detecting and combining correlation outputs or series from multiple adjacent data blocks.
MS: Mobile Station, such as a cell phone that has a baseband modem for communicating with one or more base stations. MS's referenced in this disclosure include a GPS receiver to provide position determination capabilities.
PDE: Position Determination Entity. A system resource (e.g., a server) typically within the CDMA network, working in conjunction with one or more GPS reference receivers, which is capable of exchanging GPSrelated information with an MS. In an MSAssisted AGPS session, the PDE can send GPS assistance data to the MS to enhance the signal acquisition process. The MS can return information such as pseudorange measurements back to the PDE, which is then capable of computing the position of the MS. In an MSBased AGPS session, the MS can send computed position results to the PDE.
Pseudorange measurement: A process employed by GPS receivers and based on signal processing techniques to determine the distance estimate between the receiver and a selected satellite. The distance is often measured in terms of signal transmission time from the SV to the receiver. The “pseudo” refers to the fact that the clock of the receiver is typically in error and therefore the measurement contains an uncompensated clock error term. In some cases the transmission time of the GPS signal is also unknown, or ambiguous, which provides a further timing error. In some cases pseudorange measurements are expressed in meters, by simply multiplying the transmission time by the speed of light. The units of measurement (e.g. meters of seconds) are usually clear by the context of the discussion.
SV: Satellite Vehicle. One major element of the Global Positioning System is the set of SV's orbiting the Earth and broadcasting uniquely identifiable signals.
UMTS: Universal Mobile Telephone Service: A third generation cellular standard utilizing a form of CDMA, designed to be the successor to GSM.
Table of Variables  
Variable  Description 
a  scale shift or stretch factor for time parameter. 
A  amplitude of a GPS signal. 
d(t)  a data sequence; for GPS, the data sequence biphase modulates the carrier 
frequency f_{0 }at 50 baud.  
ε  small frequency offset (e.g. ε = 50 Hz). 
f  ideal carrier frequency transmitted from the SVs. 
F(t)  waveforms corresponding to one frame of a carrier modulated by a PN sequence 
that is unique for each SV. F(t) is repeated once each “PN frame.” F(t) is also  
referred to as one frame of a “PN sequence”, although strictly speaking it is a  
waveform that is modulated according to a frame of a PN sequence.  
f  received carrier frequency; the carrier frequency received by the MS's is 
typically within a range of frequencies (Δf) about the ideal frequency f_{0}.  
f_{e}  residual carrier frequency: the remaining frequency after conversion of received 
carrier frequency to near baseband.  
f_{r}  frame rate of a waveform modulated according to a PN sequence (for C/A code 
of U.S. GPS signal, the frame rate is one kHz).  
K  number of PN frame repetitions in a data block. 
m  frequency index for a Discrete Fourier Transform (DFT) operation. 
n  index to a sequence of integers, typically representing a time sample number. 
N  number of discrete samples in an entire data block (N = KR). 
P(t)  waveform of a repeating PN sequence (i.e., P(t) includes K periodic repetitions 
of F(t)). In one embodiment the transmission (e.g., chip) rate is 1.023 MHz, F(t)  
has a length of 1023 chips, and hence the PN frame rate would be 1 kHz, and  
P(t) has length K × 1023 chips. In some cases P(t) is referred to as a repeating  
PN sequence, although strictly speaking it is a waveform modulated in  
accordance with a repeating PN sequence.  
r  output time variable. 
R  number of discrete samples in one PN frame period. 
s(t)  a received GPS signal in absence of noise. 
s_{b}(t)  received GPS signal, frequency converted to near baseband. 
S  sampling rate. 
T_{c}  time period over which a block of data is taken for coherent processing; also 
referred to as the “data block period”, or the coherent integration (processing)  
time of a signal.  
T_{s}  sample time period, i.e. T_{s }= 1/S. 
T_{r}  period of waveform modulated according to a PN sequence, i.e. T_{r }= 1/f_{r}. 
x(at)  time waveform received shifted by scale shift (or “stretch”) factor a. 
x(n)  a set of data samples for a data block: x(0), x(T_{s}), x(2T_{s}), x(3T_{s}), . . . , x(NT_{s}). 
y(f/a)  the Fourier transform of x, shifted by scale shift (or “strech”) a. 
y(m)  the FFT (or DFT) of x(n) evaluated at frequency index m. 
z(r)  matched filter output as a function of output time variable r. 
θ  the phase of the carrier frequency waveform. 
The terms correlation, convolution and matched filtering may be utilized herein. The term “correlation” is used in its broadest sense to include any type of comparison, such as coherent integration with or without incoherent integration. For purpose of illustration herein, “coherent” correlation may be used to refer to correlation processes using only coherent integration For example, coherent correlation, as applied to two series of numbers, typically means the term by term multiplication of corresponding members of the two series followed by the integration, or summation, of the series. This is sometimes referred to as “serial correlation” and results in an output that is a single number. “Incoherent” correlation may be used to refer to processes that include incoherently (i.e. nonlinearly detecting and) adding the results of several adjacent coherent integration processes.
The term “convolution” as applied to two series of numbers is the same as that commonly used in the art and is equivalent to a filtering of the second series of length p with a filter, corresponding to the first series, having an impulse response of length q. The result is a third series of length p+q−1. The term “matched filtering” refers to a convolution, or filtering, operation in which the aforementioned filter has an impulse response that is the timereversed complex conjugate of the first series. The term “fast convolution” is utilized to indicate a series of algorithms that computes the convolution operation in an efficient manner. Fast convolution algorithms often employ FFT algorithms, but are not required to do so.
GPS Environment
Reference is now made to
A user 13, on foot, is illustrated grasping the MS 14; the user may be standing, walking, traveling in a car, or on public transportation, for example. It should be apparent that the mobile station may be positioned in wide variety of environments, and may be stationary or moving.
The GPS satellites (SV's) 11 comprise any group of satellites broadcasting signals that are utilized for positioning a GPS receiver. Particularly, the satellites are synchronized to send wireless signals 12 phased to GPS time. These signals are generated at a predetermined frequency, and in a predetermined format. In a current GPS implementation, each SV transmits a civilian type of GPS signal on the L1frequency band (at 1575.42 MHz) in a format that is in accordance with GPS standards. As discussed in the background, when the GPS signals are detected by a conventional GPS receiver in the MS, the GPS system attempts to calculate the amount of time elapsed from transmission of the GPS signal until reception at the MS. In other words, the GPS system calculates the time required for each of the GPS signals to travel from their respective satellites to the GPS receiver. The pseudo range (in meters) is defined as: c·(T_{user}−T_{sv})+cT_{bias}, where c is the speed of light, T_{user }is the GPS time when the signal from a given SV is received, T_{sv }is the GPS time when the satellite transmitted the signal and T_{bias }is an error in the local user's clock, normally present in the GPS receiver. Sometimes pseudorange is defined with the constant “c” omitted. In the general case, the receiver needs to resolve four unknowns: X, Y, Z (the coordinates of the receiver antenna), and T_{bias}. For this general case, resolving the four unknowns usually requires measurements from four different SV's; however, under certain circumstances, this constraint can be relaxed. For example, if an accurate altitude estimate is available, then the number of SV's required can be reduced from four to three. In socalled assisted GPS operation, T_{sv }is not necessarily available to the receiver and instead of processing true pseudoranges, the receiver relies primarily upon code phases. In a current GPS implementation, the code phases have onemillisecond time ambiguities, since the PN codes repeat every one millisecond. Sometimes the data bit boundaries may be ascertained, thus producing only 20millisecond ambiguities.
The base stations 10 comprise any collection of base stations utilized as part of a communication network that communicates with the MS 14 using wireless signals 20. The base stations are connected to a cellular infrastructure network 15 that provides communication services with a plurality of other communication networks such as a public phone system 16, computer networks 17 such as the Internet, a position determination entity (PDE) 18 (defined above), and a variety of other communication systems shown collectively in block 15A. A GPS reference receiver (or receivers) 19, which may be in or near the base stations 10, or in any other suitable location, communicates with the PDE 18 to provide useful information in determining position, such as SV position (ephemeris) information.
The groundbased cellular infrastructure network 15 typically provides communication services that allow the user of a cell phone to connect to another phone over the phone system 16; however the base stations could also be utilized to communicate with other devices and/or for other communication purposes, such as an internet connection with a handheld personal digital assistant (PDA). In one embodiment, the base stations 10 are part of a GSM communication network; however, in other embodiments other types of synchronous (e.g., CDMA2000) or asynchronous communication networks may be used.
A GPS position location system 27 in the MS is connected to a GPS antenna 28 to receive GPS signals 12 that are transmitted at or near the ideal GPS frequency. The GPS system 27 comprises a GPS receiver 29 that includes frequency translation circuitry and an analogtodigital converter, a GPS clock, control logic to control the desired functions of the GPS receiver, and any suitable hardware and software for receiving and processing GPS signals and for performing any calculations necessary to determine position using any suitable position location algorithm. In the illustrated embodiment, the analog to digital converter is connected to the buffer memory in the position location system, and a buffer memory is coupled to the DFT circuitry to provide and store the data during the DFT operation. In some assisted GPS implementations the final position location calculations (e.g., latitude and longitude) are performed at a remote server, based upon code phases and other information sent from the GPS receiver to the remote server. Some examples of GPS systems are disclosed in U.S. Pat. Nos. 5,841,396, 6,002,363, and 6,421,002, by Norman F. Krasner. The GPS clock is intended to maintain accurate GPS time; however, since most often accurate time is not available before a location fix, it is common practice to maintain time in the GPS clock software by its estimated value and an uncertainty associated with that value. It may be noted that after an accurate GPS location fix, the GPS time will often be very accurately known, (within a few tens of nanoseconds uncertainty in the current GPS implementations). However, when the final position location calculation is done at a remote server, this accurate time may only be available at the server.
A mobile device control system 25 is connected to both the twoway communication system 22 and the position location system 27. The mobile device control system 25 includes any appropriate structure, such as one or more microprocessors, memory, other hardware, firmware, and software to provide appropriate control functions for the systems to which it is connected. It should be apparent that the processing steps described herein are implemented in any suitable manner using one or more of hardware, software, and firmware, subject to control by the microprocessor.
The control system 25 is also connected to a user interface 26, which includes any suitable components to interface with the user, such as a keypad, a microphone/speaker for voice communication services, and a display such as a backlit LCD display. The mobile device control system 25 and user interface 26, connected to the position location system 27, provide suitable inputoutput functions for the GPS receiver and the twoway communication system, such as controlling user input and displaying results.
Coherent Processing
The coherent processing method is described with reference to
Observe GPS Signal (30)
In
s(t)=A d(t)P(t)exp(j2πft+θ) (A1)
where A is the signal amplitude, d(t) is a relatively low rate (e.g., 50 baud) data sequence that modulates the carrier (e.g., by biphase modulation), P(t) is the waveform consisting of a repeating set of frames of a PN sequence F(t), f is the carrier frequency (which is ideally equal to f_{0}), and θ is the carrier phase. It may be noted that equation (A1) is a complex representation of the carrier, which can be useful if quadrature sampling methods are utilized to process the signal; of course other representations may be used as appropriate. In real world situations it should be recognized that the various parameters are not completely stable, but for the explanatory purposes we assume that the signal amplitude and various modulation rates are approximately constant.
Each GPS satellite (SV) transmits a unique PN waveform F(t) shown at 46, which is a series of symbols (chips) transmitted at a predetermined rate. The PN waveforms are distinguished from one another by the particular PN sequence used to biphase modulate the carrier. For example, these sequences are chosen from a set of Gold codes, in C/A waveforms of the U.S. GPS system. Again, in the following we sometime use the terminology “PN sequence” or “PN frame” for F(t), which is not strictly correct, since the PN sequence is actually the sequence of numbers that is used to construct the modulating signal of a carrier, thus producing the waveform F(t). However, it will be clear from the context that “PN sequence” used in this manner means a waveform modulated by the PN sequence.
In one example, the chip rate is 1.023 MHz and hence the PN frame rate would be about 1 kHz. This waveform F(t) is repeated continually; for example a first code from a first satellite SV_{1 }repeatedly transmits the unique sequence F_{1}(t), SV_{2 }repeatedly transmits the unique PN sequence F_{2}(t), and so forth. The GPS receiver is programmed with the unique PN sequences for all GPS satellites that may be inview. These PN sequences may be used in an algorithm to identify the particular satellite; particularly, when a satellite signal is received in a GPS receiver, the PN sequence is utilized to identify the satellite that transmitted the received signal. However, initially, the GPS receiver does not know the actual received code phase epoch, which, as described above, may range over a full PN frame (e.g., a period of one millisecond or 1023 chips). Furthermore, the receiver does not know if the GPS signal associated with a particular PN code is detectable, since it may be significantly attenuated by various obstructions and/or the particular SV may not be inview. Hence the receiver must search, either serially or in parallel, over the epoch uncertainty range in an attempt to detect the hypothesized signal and align the epoch of a received GPS frame with that of a locallygenerated reference frame.
It should be noted that, in an actual GPS environment, the GPS receiver simultaneously receives a multiplicity of signals like the theoretical signal specified in equation (A1), each having a unique PN sequence F(t). For example, in a typical situation, the GPS receiver typically receives eight to twelve signals from a variety of inview satellites at any time, and the various parameters differ from one another due to differing path lengths, directions of arrival, and Doppler frequency shifts, for example. For purposes of illustration, the following disclosure first discusses processing one of the signals of the theoretical form of equation (A1), and then later it is shown how the processing algorithms described herein may be advantageously used to process multiple signals, each having a theoretical form of equation (A1).
It may also be noted that, when the GPS signals reach the receiver, they are often heavily corrupted by additive noise, and perhaps also corrupted by other noise or interference. In addition, as previously mentioned, the carrier frequency and chip rate may appear to be shifted slightly from its original value, primarily by Doppler effects. In other words, the carrier frequency may shift slightly as observed by the receiver in the MS due to motion of the SV and that of the MS, and therefore when the receiver receives the signal, the actuallyreceived carrier frequency may vary from its ideal predetermined carrier frequency f_{0 }by an amount called the “residual frequency”. In addition, errors in the MS local oscillator also cause the carrier frequency to vary from its ideal frequency.
Remove Carrier Frequency from GPS Signal (31)
Referring again to
Typically, the residual frequency arises primarily due to Doppler effects. In addition, the receiver itself may introduce a slight frequency shift during processing the signal. The sum of these two errors from the ideal carrier frequency may be represented by a certain maximum tolerance (Δf); therefore the actual received carrier frequency is typically within the range of f_{0}±Δf. The residual frequency f_{e}, which is equal to the frequency remaining after the receiver has attempted to reduce the original carrier to zero, is typically in the range of several hundred hertz to several kHz, although the residual frequency may be more or less in any particular set of circumstances.
In assisted GPS systems, a predicted Doppler correction for all GPS signals is transmitted (in one form or another) from a PDE to the GPS receiver, and a list of the GPS satellites that may be inview is also sent to the receiver so that the GPS receiver can more efficiently search for satellite signals. A predicted data stream can also be provided by the PDE.
In its memory, the receiver has stored the PN codes (or a representation thereof—e.g. the DFT of these codes) corresponding to all the GPS satellites that may be in its view.
Digitize and Store Processed GPS Signal (32)
At 32, the processed GPS signal is digitized (i.e., sampled) over a predetermined time period in the analogtodigital converter (if it has not been previously converted), and then stored in the buffer memory in the GPS receiver. There is no theoretical restriction on the size of the data set, or the sample rate of the data, although it is sometimes beneficial that the sample rate be a multiple of 1.024 MHz and that the data set size be a multiple of 1024. Therefore the signal of equation (A2) is considered to be a sampled signal, where the sample rate may be advantageously set to a number that is 1.024 MHz or 2.048 MHz, so that 1024 or 2048 samples occur over the PN frame period of one millisecond. Note that due to Dopplerinduced errors, this sample rate is not quite equal to the chip rate or twice the chip rate. One reason for choosing this sample rate is that if sampling is performed at 1024 or 2048 MHz, then the resulting number of samples, over a one millisecond frame period, is a power of two, which is convenient for efficient FFT processing. That is, one frame of data is a multiple of 1024 samples, a convenient size for efficient FFT's, and the total data set size (for coherent processing) is also a multiple of 1024, and preferably a power of two times this length. This restriction is not essential, however, since efficient algorithms still exist when the number of samples per frame is not a power of two.
Select Data Block (33)
At 33, a data block for coherent processing is defined by selecting a portion of the stored digital data over a predetermined coherent processing period T_{c}. The time period over which the data is combined for coherent processing is typically chosen to include a significant, integral number of PN frames (e.g., 20 PN frames); however, the coherent processing block should not be chosen to be too long; particularly, it may be disadvantageous to try to coherently process extremely large blocks (e.g., a 1 second block) since the stability of the residual carrier frequency f_{e }and other multipath effects (and possibly other factors) over that long time period may restrict or prevent any improvement in performance. As will be discussed, it is advantageous to choose T_{c }to be an exact multiple of one PN frame period T_{r}.
Referring briefly to
Removal of Data Sequence (34)
At 34, the data sequence is optionally removed. After the data sequence d(t) has been removed and the theoretical signal of equation (A1) has been converted to a frequency near baseband the remaining signal (ignoring noise and interference) has the form:
s _{b}(t)=A P(t)exp(j2πf _{e} t+θ) (A2)
where f_{e }is the residual frequency after conversion of the carrier frequency to near baseband.
Although optional, removal of the data sequences d(t) before processing can be very useful. To assist in data sequence removal, in some assisted GPS systems the predicted data sequences d(t) are sent (e.g., from a server) to the GPS receiver, together with some approximate timesofarrivals of the GPS signals. In these cases, the GPS receiver can remove the data sequence d(t) and hence remove the pseudorandom polarity inversions that may occur every twenty milliseconds in the signal of equation (A1) due to the data sequence d(t). By removing the random polarity inversions (i.e., by removing d(t)), the coherent integration time can be increased to longer time intervals than one data bit period, for example greater than 100 milliseconds. Advantageously, increasing coherent integration time can improve the sensitivity of the GPS acquisition process. As indicated previously, some future modes of GPS may contain signaling components containing no data. In these situations the coherent integration period is not limited to a data bit period.
Perform Fourier Transform on Data Block (35)
Referring again to
At 35, the data block is coherently processed using a Fourier Transform process. This step may be called the “forward transform” process. For example, a Fast Fourier Transform (e.g., FFT or DFT) of the signal s_{b}(t), sampled over the time period T_{c}, may be performed:
y(f)=FFT(s _{b}(t) from t=0 to t=T _{c}) (A3)
The forward transform process can be performed in a variety of ways. One wellknown approach is decimation in time; another approach is decimation in frequency. Other fast algorithms may be employed as appropriate or useful, such as chirp ztransforms or number theoretic transforms.
The FFT of an arbitrary signal (shown in
Particularly, the comb spectrum shown in
It may be noted that the separation of the strong spectral lines of the comb, measured in Hz, is equal to the frame rate f_{r}. Measured in frequency index difference, it is K indices, i.e., the number of frames in the coherent data block.
While
Reference is now briefly made to
Unlike
Select Satellite that may be in View, Hypothesize Initial Frequency Offset (36)
Returning to
At 36 a, a particular satellite that may be in view is selected or “guessed”. The selection of any particular satellite may be random, or preferably may be based upon any suitable information, such as history or a list provided by the PDE. As will be discussed, the PN code for the selected satellite will be tested over all frequency hypotheses (in a range typically determined by the receiver), at least until a match is found or all hypotheses have been exhausted, and then at 36 c, the next satellite will be selected and the corresponding PN code is tested over all frequency hypotheses, and so on until all candidate satellites have been selected, or until signals from a sufficient number of satellites have been found to complete a location fix.
Also at 36 a, an initial hypothesis is made as to the residual frequency. If sufficient information is available to the GPS receiver, (e.g., a prior location fix has been made or estimated Doppler corrections are available), then this initial hypothesis and subsequent hypotheses may be made based upon this information. If no information is available, then a best guess may be made and searching would begin.
Provide FFT of GPS Code of Selected Satellite (37)
Returning again to
B(f)=FFT(P(t)) from t=0 to t=KT _{r} =T _{c}) (A4)
The result will be a comb spectrum including a series of evenlyspaced lines like the example shown in
However, it is likely to be more efficient to precompute the Fourier transform of the repeated sequence, P(t) and then store only the nonzero Fouriertransformed values for each GPS code to allow quick use whenever required. It is easy to see that these nonzero values may be derived from the Fourier transform of F(t), rather than P(t), a fact that may reduce the computational burden. That is, it is normally sufficient to compute the FFT of only one repetition of F(t), rather than K repetitions, as was indicated in (A4), since the FFT of the repeated sequence may be derived from this shorter FFT. It should also be clear that the reference GPS code is normally assumed to have a code phase offset of zero and a carrier frequency offset of zero, and hence should be centered at 0.0 Hz and should look like the graphical depiction of
A subtlety in the above computation concerns the fact that the U.S. GPS PN code is of length 1023 and a preferred FFT size is a power of two, typically either 1024 or 2048 in this discussion. If the FFT is precomputed, then corresponding to a sample rate of 1.023 MHz, an appropriate procedure to create the FFT of appropriate size, would be to perform a 1023 point FFT of the reference and append an extra zerovalued sample between indices 512 and 513. Similarly, corresponding to a sample rate of 2.046 MHz, an appropriate procedure to create the FFT of appropriate size, would be to perform a 2046 point FFT of the reference PN (sampled at two samples per chip) and append two extra zerovalued samples between indices 1024 and 1025. These procedures are interpolation techniques performed in the frequency domain, and are more efficient in computation than performing an equivalent resampling approach in the time domain. In either case, the FFT of a repeated reference sequence can then be computed by simply inserting an appropriate number of zerovalued samples between each of the reference frequency samples corresponding to the FFT of one PN frame.
Select (“Prune”) a Subset of the Data Frequency Samples (38)
When the data frequency samples were originally calculated in the FFT process at block 35, the residual frequency was unknown. In order to accurately and efficiently acquire a GPS signal, this unknown residual frequency must be found. In order to determine the residual frequency, a “trial and error” process may be utilized in which a series of residual frequencies are hypothesized, calculations are performed for each hypothesis, and the results are analyzed to search for a match. It should be recognized that the number of hypotheses may be large, and processing time increases with the number of hypotheses tested.
At 38, a subset of the data frequency samples is selected, i.e., “pruned”, responsive to a hypothesized residual frequency. As shown in
For example, referring again to
If the sampling rate is 1.024 MHz, the block size is 20 msec, and there are 20 PN sequences in the block, then there are only 1024 lines of the DFT of P(t) with appreciable energy, since the spacing of the adjacent comb lines of the received signal is 1 kHz. This comb spacing limits the subset to only 1024 data frequency lines, and therefore a correspondingly reduced size inverse FFT can be utilized in subsequent processing.
As another example, if the sampling rate is 2.048 MHz, then there would be 2048 nonzero value comb lines, again with a 1.0 kHz comb frequency spacing, but now the energy extends over the larger 2.048 MHz passband. It may be noted that it is not necessary to sample at a rate that is a multiple of the frequency separation (e.g., 1.0 kHz), nor is it necessary that the sample rate be a power of two times 1.0 kHz; a comb spectrum for s_{b}(t) would still remain. It is highly desirable, however, that the total sample period T_{c }be a multiple of one millisecond in order to achieve true periodic convolution. Even this requirement can be dispensed with, as discussed later, likely at the cost of some performance or speed degradation.
Reference is now made to
The terminology “pruning” refers to the fact that we are selecting only every Kth sample from the frequency data. In the prior example where T_{c }equaled 20 PN frames, K equaled 20, i.e., we need only select every 20^{th }sample of the FFT data for use in subsequent processing. More generally, K is the number of repetitions of the PN code in the coherent data block being processed. Such pruning leads to dramatic reduction in the amount of subsequent processing.
Reference is now made to
In
In summary, the information useful for identifying a GPS signal's presence is contained substantially only within spectral lines that are displaced from one another by a constant amount (1 kHz in this example) and offset by the residual frequency; therefore following a hypothesis of a residual frequency, the set of spectral lines (the comb) corresponding to that frequency offset can be selected from the FFT and the rest ignored for subsequent matched filter calculation purposes corresponding to the hypothesized residual frequency. Advantageously, this reduced number of spectral lines can significantly reduce the number of subsequent calculations required and thus reduce processing time for each hypothesized residual frequency; particularly, if we are utilizing a sample rate of S, instead of having to perform an inverse FFT of size ST_{c}, as would be otherwise necessary in the matched filter operation of step 39, we need only perform an inverse FFT of size S/1 kHz. As an example if we assume that T_{c}=128 milliseconds, we might normally have to perform an inverse FFT of size 128×1024, if the sample rate were 1.024 MHz. Taking advantage of the sparseness of the spectrum (i.e., the fact that the GPS signal has a comb spectrum) we need now only compute an inverse FFT of size 1.024 MHz/1 kHz(i.e., 1024), a processing savings well exceeding a factor of 128 (more precisely: 1.7×128). Furthermore, the processing savings advantage improves as the total coherent processing time T_{c }increases; particularly it can be seen that the FFT size reduction is related to the number of repetitions of the PN sequence F(t), i.e., the FFT size reduction factor is improved with a greater number of PN frames coherently integrated.
Compute a Correlation Series (39)
At 39, an operation is performed to form a correlation series from the subset of data frequency samples and the reference frequency samples (e.g., GPS code). To accomplish this an FFT based matched filter operation may be performed as follows:
Multiply the selected subset of data frequencies by the complex conjugate of the FFT of the GPS code; and ______(A5)
Perform an inverse FFT of the result of equation (A5) and perform a detection operation upon this resulting data set. ______(A6)
The result will be a circular convolution of s_{b}(t) and P(t), which will provide the proper correlation information, assuming that the duration of s_{b}(t) is an integral number of PN frames. Note that this basic procedure required processing long data sets of duration T_{c}, i.e., this procedure required performing large size forward FFT's. However, there are efficient ways of performing such large FFT's, as are well known. The computational benefits accrue from having to perform only very small size inverse FFT's, due to the pruning procedure. Since many inverse FFT's may need to be performed corresponding to many frequency hypotheses, the resulting computational savings may be dramatic. This will be further justified mathematically in a later discussion.
For purposes of explanation, the method of steps 3339 disclosed above corresponds to processing one block of data in a coherent manner, which is a type of correlation termed herein “coherent correlation” or “coherent processing.” In order to improve sensitivity, the correlation outputs from a number of coherent correlation processes may be detected and combined over a number (e.g., 2 to 2000 blocks, typically 5 to 200 blocks) of adjacent time intervals to provide the correlation results. This process is termed “incoherent correlation”, and is later discussed in more detail together with
Search Correlation Results for Match (40)
At 40 in
Referring again to
Determining Code Phase Offset (42)
As discussed above for example with reference to
At 42, following a positive search result (i.e., after a match has been found in step 40), the code phase offset is determined from the results of the matched filter operation of step 39 Particularly, before the matched filter operation, the number of possible code offsets is known; in an example disclosed herein in
At 43, a decision is made as to whether or not signals from additional satellites are to be searched. This decision is made in accordance with any suitable criteria; for example, if signals from sufficient satellites have already been found to make a location fix, or if the list of possible inview satellites has been exhausted, then a decision may be made to stop searching and therefore, as indicated at 44, the acquisition operation is complete. However, if signals from more satellites are to be searched, then at 36 c, the next satellite is selected, an initial residual frequency is hypothesized, and steps 3742 are performed with the new assumptions.
Summary and Additional Discussion
Utilizing the knowledge that the PN sequence F(t) repeats a number of times in a coherent processing data block, as discussed herein it has been recognized that a simpler inverse FFT procedure is possible as part of the overall matched filter procedure, which advantageously reduces computation time. It may be noticed that if only one Doppler hypothesis were to be searched, the improvement in processing time would not be particularly significant. However, because searches must usually be performed over a large number of Doppler hypotheses (e.g., a search over ±500 Hz would not be unusual) then this processing savings advantage as described herein quickly becomes very substantial. One reason for the processing savings is that each Doppler hypothesis requires a separate inverse FFT to be performed; however, in the approach discussed herein, the inverse FFT size is independent of the size of the coherent frequency block due to the fact that one need only process frequency samples at the hypothesized comb frequency locations. The number of such frequency samples is easily seen to equal the number of data frequency samples over one PN frame. In the above example, with a 128millisecond processing block size, the inverse FFT sizes required are reduced by a factor of 128, resulting in improved processing speed by a factor of more than 128. Although a large forward FFT must be performed as in Step 35, this large operation need only be done once per GPS code searched, and in some cases one forward FFT may be shared for multiple hypothesized GPS codes.
Typically, to search over a large Doppler range, a correspondingly large number of Doppler hypotheses are sequentially made and performed one after the other, which thus would require performance of a large number of inverse FFT's. For example, to search over a range of residual carrier frequencies f_{e}±2 kHz, with a 128millisecond coherent integration time, a number of Doppler hypotheses would be required, i.e. a number of inverse FFT's would be performed equal to at least 512 (4000 kHz×128 msec). In the prior example the inverse FFT sizes need only be 1024 points rather than 131072, which would result in a savings of computation time by a factor of about 218 (noting that FFT processing time is proportional to N log(N), where N is the transform size). For example, using currently available technology, a 1024point FFT can be performed in under 0.5 milliseconds using low cost DSP integrated circuits, thus resulting in overall processing time for the entire set of inverse FFT's of less than 0.26 seconds, whereas without taking advantage of the sparseness of the data, the processing time would be about 1 minute. Furthermore, since one must search over a multiplicity of hypothesized GPS PN codes, the processing time required with conventional FFT processing may become impractical, whereas with the disclosed approach it is easily practical.
The searching over various Doppler hypotheses is simplified by recognizing that the adjacent spectral lines of the FFT are separated from one another by a predetermined number, which in this example is 1/T_{c }Hz (e.g., if T_{c}=128 msec, then 1/T_{c}=1/128 msec =7.813 Hz). Therefore, for a given PN code, it is not necessary to again perform the forward FFT for each frequency: to alter a frequency hypothesis, we need only shift the FFT of s_{b }by one index position (the index value is determined as appropriate to make signal acquisition likely without unnecessary effort). Let y equal the FFT of s_{b}. In the example where the sample rate is 1.024 MHz, and T=128 msec, if the frequency hypothesis were zero, we would then process samples of y numbered 0, 128, 256, . . . , etc. If the residual frequency hypothesis is 7.813 Hz, then we would process samples numbered 1, 129, 257, etc. If the residual frequency hypothesis is −7.813 Hz we would process samples 131071, 127, 255, and so forth. (Note that index 131071 is equivalent to −1, since the spectrum is periodic with period 131071). The pruned processed block for each case is multiplied by the complex conjugate of the nonzero FFT samples of the reference GPS waveform. The result is inversetransformed to provide a matched filter output representative of one PN frame. A peak magnitude (or magnitudesquared) found above a threshold in this output is representative of the presence and time of arrival of a received GPS signal having the GPS signal number and the Doppler frequency corresponding to that used in the processing sequence. As will be discussed below, in some case we might desire to shift the FFT by a fraction of an index number. This may be done using frequency interpolation methods, as discussed below, rather than by simply rotating or shifting the frequency set.
Incoherent Processing
The method of
This modification may be understood more precisely with the aid of
Examining
In 147, one branches to the right in order to repeat the processing of 138, 139, and 146 upon the next block of data, unless all blocks of data have been processed, at which point the processing flow proceeds to 140. When processing proceeds to 140 one has combined all the correlation series that are desired for determining a matched condition. The combined correlation series at this point is termed the “final correlation series.” The final correlation series is examined for a matched condition, typically a peak above a detection threshold, and the corresponding code phase offset is found, in a manner similar to that explained for
In the above description it is noted that the repetition of operations 138, 139, and 146 are upon successive blocks of data, but the hypothesized SV, reference frequency samples, and residual frequency are the same for each repetition. If no match is found in 140, then a new residual frequency is chosen in 136 b (unless the set has been completely searched) and the processing 138, 139, 146 begins anew starting with the first data block (145 reinitializes the block number). Since one has previously computed FFTs on all data blocks in step 135, it is not necessary to do any more forward FFTs when altering the hypothesized next residual frequency. That is, the frequency samples for each of the data blocks have been stored in a buffer and may be reused for each of the subsequent residual frequency hypotheses.
After a match has been found or after all residual frequencies have been exhausted the processing proceeds to 143 where, if more SVs need to be examined, one selects the next SV and initial frequency in 136c and then proceeds to step 133. In some cases, such as when no data sequence is present, one could alternatively proceed at this point to step 136 a and reuse the data frequency samples from 135 that had already been computed by prior FFT operations.
The following description is one explanation of operation of one method disclosed herein, for purpose of illustration and developing a greater understanding of the method.
First, consider the way in which an inverse FFT operation is performed. The originally sampled time data can be represented as x(n): n=0, 1, 2, . . . , which is shorthand for the data samples x(0), x(T_{s}), x(2T_{s}), . . . , where T_{s }is the sample time period. The discrete Fourier transform (“DFT”) of this sampled data is denoted by y(0, 1 ,2 , . . . ). The DFT of this data effectively denotes frequency samples at frequencies 0, 1/(NT_{s}), 2/(NT_{s}), . . . , m/(NT_{s}), . . . , where m is the sample number. The DFT y(m) is defined for each m by:
Note that the frequencies of the DFT corresponding to m>N/2, are actually negative frequencies, since by circular symmetry a frequency corresponding to index m (i.e., frequency m/(NT_{s})) is equivalent to one corresponding to index m−N (i.e., frequency (m−N)/(NT_{s})). Now, for purposes of this explanation, assume 1) that the GPS frame period corresponds to R input samples, 2) as before, any satellite data has been removed, 3) the block size N corresponds to K frames, that is, N=KR, and 4) any Doppler effects on the signal modulation are negligible. These assumptions allow the FFT algorithm to perform a periodic convolution.
We are only interested in finding R samples out of a matched filter operation, since the matched filter operation is essentially the circular convolution of the signal data with the periodically repeated reference, as indicated previously. Hence, the matched filter result will also be periodic with period R. Under these circumstances we can provide the matched filter output by operating upon y(m) of equation (B1) in the wellknown manner:
where g is the FFT of the GPS reference PN waveform [sampled at the same rate as x(n)], repeated K times, the asterisk represents the complex conjugate, and r is the output time variable, which need only range over [0, 1, . . . , R−1]. In equation (B2) we are hypothesizing that the residual carrier frequency of the signal y(m) is zero. As indicated before, because the PN sequence is periodic every frame, i.e., every R samples, the function g(m) has nonzero values every N/R=(KR/R)=K samples (in frequency). For example, if N corresponds to 20 frames of GPS data, then only every 20^{th }sample of the FFT (beginning with the first) of g is nonzero. Accordingly, the product within the summation of equation (B2) is nonzero only for every 20^{th }sample, and hence we can write (B2) as:
The last summation is the Rpoint inverse DFT. Hence, it has been shown that the inverse DFT required for the matched filtering operation may be performed using only an Rsample FFT algorithm, which advantageously reduces processing time and memory requirements. Furthermore only an Rpoint inverse DFT is necessary, no matter how many PN frames of data, K, are being processed, as long as the abovementioned conditions are met. Note that equation (B3) is mathematically identical to that which would have been obtained if the full Npoint inverse FFT were performed as in equation (B2). Note also that equation (B3) clearly shows the selection of every K^{th }point from the FFT of y for performance of the inverse FFT. This is the basis for the “pruning” procedure, that is, the selection of a subset of points for performing the inverse FFT. Equation (B3) holds whether or not the hypothesized residual carrier frequency error is correct. However, this process will only produce a strong detection indication when the residual carrier frequency error is small compared to 1/T_{c}.
The above equation (B3) corresponds to processing the transformed data samples, assuming a residual carrier Doppler shift of zero. It will produce a strong correlation peak only when the residual frequency is close to zero. In order to change this assumption, the Doppler shift is assumed to be d/(NT_{s}), where d is an integer, and equation (B3) is modified as follows:
where [ ]_{mod N }is the bracketed quantity, modulo N. Essentially, we are frequency shifting the input signal so that it has nearzero (i.e. much less than 1/Tc) residual frequency, assuming our Doppler hypothesis is correct. Equation (B4) takes advantage of the circular nature of y. Note that this modification is simply a frequency translation of y by d spectral lines, and is straightforwardly implemented by indexing the sequence y (in a circular manner), beginning d positions relative to the first element of y. This approach eliminates the prior art restrictions, discussed in the background section, which would otherwise effectively limit searches over a range greater than about −500 to 500 Hz. The only restriction on the Doppler hypothesis d is an implicit one relating to a stretching of y due to timeDoppler effects (i.e., Doppler on the signal's modulation). This restriction may be removed as discussed below.
One convenient aspect of equation (B4) is that, to process a different GPS code, it may not be necessary to perform another forward transform. In some circumstances, the appropriate GPS code for “g” (e.g., the appropriate Gold Code) can be substituted into the above, and the prior transformed data can continue to be used. This can be done if the satellite data information (message) present in more than one concurrently received GPS signal is substantially the same. This condition would allow the concurrent removal of the data transmissions on the simultaneously received signals. This is possible if two conditions are met: (A) the differential ranges from the satellites are fairly small (e.g. within 300 km) and (B) the message data information is similar between SV transmissions. Item (B) often occurs—e.g., when satellite Almanac is transmitted. Also, item B is not so significant if coherent integration times are less than 20 milliseconds. It may be noted that, in GPS modes that do not contain data such as those proposed for future implementation, condition (B) would not apply and this modification may be more generally done.
Modulation Doppler Effects.
In the above explanation, the effects of Doppler shifts (including the receiver reference local oscillatorinduced “Doppler”) have been assumed to mainly affect the carrier frequency. However, if the coherent integration time NT_{s }becomes sufficiently large, then it may not be possible to ignore the effects of Doppler upon the signal's modulation (i.e., upon the PN sequence P(t)). For present purposes this modulation Doppler effect, or “time Doppler” effect, primarily alters the modulation rate, in effect “stretching” or “compressing” the signal waveform relative to the reference produced in the GPS receiver.
For example, for processing the C/A code for Standard Position Service (civilian service) on GPS, the ratio of the carrier frequency to the chip modulation rate is about 1575.42e6/1.023e6=1540. Hence, a Doppler shift of about 5000 Hz upon the carrier would result in a Doppler shift of about 5000/1540=3.25 Hz on the modulation. For processing relatively short blocks of data coherently (e.g., 20 milliseconds), such a time Doppler may not be significant. But when processing long blocks of data, the effect can greatly degrade the system sensitivity by significantly reducing the magnitude of the matched filter peak output. As a rule of thumb, if the modulation Doppler (including local oscillator effects) is p Hz, and the total block size N corresponds to T_{c }seconds, then without additional processing, it is advantageous to keep the quantity pT_{c }below about ½ in order to reduce deleterious effects.
Consider the above case in which a 10,000 Hz Doppler shift on the carrier results in a 7.143 Hz Doppler shift on the PN modulation. If the coherent block size is about 100 milliseconds, then pT_{c}=0.7143, and some degradation in system performance would be noticeable. Furthermore, the time of the peak output from the matched filter would be displaced by pT_{c}/2 chips, relative to the zeroDoppler case. It is thus clear that a large Doppler search range and a long coherent integration time will result in significant losses from time Doppler effects, if left uncorrected. This problem is particularly amplified in two important situations:
(1) Large differences between the Doppler shifts from one GPS satellite signal to another, as observed by the GPS receiver. This item has already been discussed above.
(2) An effective Doppler shift due to errors of the GPS local oscillator's frequency relative to its ideal frequency.
Regarding item (2), the GPS local oscillator may differ significantly from the ideal GPS frequency. For example, sometimes a GPS receiver can derive its local oscillator frequency from that of a synchronized cell phone, and hence achieve very low errors. However, in some situations this may not be possible. Even good temperaturecompensated crystal oscillators may have frequency errors at the GPS frequency (1575.42 MHz) of over ±3000 Hz. Although such frequency errors are not true Doppler shifts, they create both carrier and modulation shifts at the GPS receiver, similar to the Doppler shifts observed from a moving platform. Such frequency errors are common to all GPS receivers, and therefore affect all GPS signals that are processed to some extent. Nevertheless, these frequency errors can result in deteriorated performance, particularly for long coherent block sizes.
One way of coping with the above problem is to resample the input data sequence at a rate commensurate with the Doppler hypothesis of the GPS SV (satellite vehicle) and/or that due to the local oscillator errors. By resampling the signal, utilizing digital signal processing methods, the input signal can in effect be stretched or compressed so that there are again an integral number of PN frames of GPS data within the coherent processing block. Without such resampling the number of such frames in a coherent block will no longer be an integer but be either more or less by an amount that can be as large as several samples, which could result in severe deterioration of the peak signal produced by the matched filter operation.
However, resampling in the time domain, as discussed above, has one significant disadvantage, in that one must resample and perform a large forward FFT for a range of frequencies, and for a given SV. The range is such that pT_{c} is less than around ½, as indicated earlier. Unfortunately, this requirement to perform multiple forward FFT's can result in both a substantial increase of system memory requirements, and a substantial increase in processing time.
However, the above disadvantages are eliminated, and particularly the requirement to perform additional forward FFT's can be eliminated, by performing the resampling function in the frequency domain. In other words, the resampling function can be performed upon the transformed signal y, rather than in the time domain. This approach would completely avoid the requirement to perform additional forward FFT's; however, depending upon implementation, some additional storage may be needed.
To understand the basic principle behind the frequency domain resampling, one need only look at the Fourier transform relationship:
where x is the time waveform, y is the Fourier transform of x, and a is a scale shift or stretch. Clearly then, we can perform the stretching in either domain.
The stretching or compression includes resampling the frequency samples, a process that involves fractional resampling methods. From (B5) we see that if the frequency samples are called y(m) and thus these samples are initially provided at frequencies m=[0, 1, 2, . . . ]/(NT_{s}), then these samples are replaced by samples evaluated at the frequencies m/a; i.e., by samples evaluated at frequencies of m_{r}=[0, 1, 2, . . . ]/(aNT_{s})=[0, 1/a, 2/a, . . . ].
It may be noticed that this last result is only correct for positive frequencies since we must ensure that the data samples are spaced symmetrically about 0 Hz. To do this, if we reorder the initial set in the order: −N/2−1, −N/2, . . . −1,0,1 . . . , N/2−1, n/2, then the resampled set is resampled at frequencies:
[(−N/2−1)/a, (−N/2)/a, . . . , −2/a, −1/a, 0, 1/a, 2/a, . . . , (N/2)/a]/(NT_{s}) (B6)
that is, if we use the original order, it is resampled at frequencies:
m/a: for m=0, 1, 2, . . . , N/2 (B7)
N+(m−N)/a: for m=N/2+1, N/2+2, . . . , N−1 (B8)
where we note the circular nature of frequency so that frequency index m is the same as m+N or m−N.
The resampling of equation (B6) or (B7) requires evaluating the frequency response at frequencies that are “between” the normal discrete frequencies evaluated by the DFT. But this is relatively easy to do with a “sinc” interpolator, for example. Since the input data is time limited, we can evaluate the (complex) frequency response at frequency ε, ε<0.5 Hz, relative to a set of spectral lines, through a convolution procedure. For example, to evaluate the spectral response at frequency y(m_{0}+ε), where m_{0 }is an integer, we form the product:
where m ranges over all possible values (i.e., from m−N/2+1 to m+N/2).
A simple approximation to this calculation would require only two or three values of m. An estimate of the loss due to the twoterm evaluation of equation (B9) indicates that such a sensitivity loss is less than 1 dB, if ε ranges over the range from −0.5 to 0.5 Hz. For most modulation Doppler shifts of interest, the stretching due to equation (B5) may be considered to be fairly constant for a relatively large number of consecutive frequency samples. Hence, the interpolation procedure of equation (B9) can use the same values for the sinc weighting coefficients in order to determine a large number of consecutive resampled spectral values.
The resampling method described above thus allows use of an algorithm in which the frequency data y is broken into a series of smaller blocks, for example of size 1024 each, and each block is resampled using an interpolation procedure with a fixed set of coefficients. Prior to processing a block, the coefficients are either computed or are looked up in a table. This procedure could greatly reduce the processing burden for the resampling operation. For example, if a twopoint interpolation procedure like equation (B9) were employed, the resampling procedure would require only four real multiplies and two additions to compute each interpolated value (ignoring the abovementioned table lookup). This method can be compared with, for example eight butterflies per data sample necessary to compute an FFT on a block size equal to 64K. These butterflies would require 32 real multiplies and 48 additions, a computational increase by a factor of about 16 relative to the interpolation in the frequency domain. Hence frequency domain resampling is believed to be much more practical and efficient than timedomain resampling.
As disclosed herein, resampling is useful to compensate for modulation Doppler when either processing very large ranges of Doppler shifts, and/or when processing signals from different SV's. In such cases the same Fouriertransformed data set can be utilized and therefore it would be unnecessary to deal with the original time data. As indicated before, however, processing different SV's with the same Fouriertransformed data set may be limited to situations in which the satellite message data is similar, in order to allow its removal prior to the initial coherent processing. In any case it is useful to retain the original Fouriertransformed data set even after a resampling operation is performed, in the event that a second and additional resamplings are required. If the original Fouriertransformed data set is not available, it would be necessary to perform resampling on a resampled set, an approach that could result in cumulative errors unless very precise resampling were to be performed.
The pruning operation has been defined as a selection of a subset of the frequency data from the forward FFT, due to the sparseness of the spectrum associated with the repeated PN signal—i.e. the comb line shape. When spectral interpolation is required, as discussed above, we construct, rather than merely select, the subset by interpolation between frequency samples. Nevertheless, the size of the subset so constructed is similar to the case in which simple selection is made. That is, it is typically equal to the number of signal samples per one PN frame—e.g., in the prior examples it was either 1024 or 2048 samples, corresponding to sample rates 1.024 MHz or 2.048 MHz. The inverse FFT sizes are accordingly these sizes as well. Consequently the definition of “pruning” extends to the construction of a subset of frequency samples via an interpolation procedure, as well as the direct selection of a subset of frequency samples.
In a similar manner the interpolation procedure may be utilized when one wishes to alter successive frequency hypotheses by increments that are smaller than an FFT line spacing—for example, an increment of onehalf a line spacing may be desired. Again, the pruning definition extends to the construction of a subset of frequency samples via an interpolation procedure in which the frequency hypothesis is altered by a fraction of the FFT line spacing.
It will be appreciated by those skilled in the art, in view of these teachings, that alternative embodiments may be easily implemented.
For example in the prior discussion, as exemplified for example by
Claims (45)
Priority Applications (2)
Application Number  Priority Date  Filing Date  Title 

US54012004 true  20040128  20040128  
US10936177 US7702002B2 (en)  20040128  20040907  Rapid acquisition methods and apparatus for GPS signals 
Applications Claiming Priority (9)
Application Number  Priority Date  Filing Date  Title 

US10936177 US7702002B2 (en)  20040128  20040907  Rapid acquisition methods and apparatus for GPS signals 
JP2006551634A JP4616282B2 (en)  20040128  20050127  Rapid acquisition method and apparatus Gps signal 
DE200560010391 DE602005010391D1 (en)  20040128  20050127  Fast procurement methods and devices for GPS signals background 
EP20050712838 EP1712010B1 (en)  20040128  20050127  Rapid acquisition methods and apparatus for gps signals background 
CA 2554412 CA2554412A1 (en)  20040128  20050127  Rapid acquisition methods and apparatus for gps signals background 
RU2006130799A RU2360263C2 (en)  20040128  20050127  Device and method of highspeed detection of gps signals 
PCT/US2005/003540 WO2005074153A1 (en)  20040128  20050127  Rapid acquisition methods and apparatus for gps signals background 
US12763026 US8111736B2 (en)  20040128  20100419  Rapid acquisition methods and apparatus for GPS signals 
US13353099 US8576895B2 (en)  20040128  20120118  Rapid acquisition methods and apparatus for GPS signals 
Related Child Applications (1)
Application Number  Title  Priority Date  Filing Date 

US12763026 Continuation US8111736B2 (en)  20040128  20100419  Rapid acquisition methods and apparatus for GPS signals 
Publications (2)
Publication Number  Publication Date 

US20050163201A1 true US20050163201A1 (en)  20050728 
US7702002B2 true US7702002B2 (en)  20100420 
Family
ID=34798950
Family Applications (3)
Application Number  Title  Priority Date  Filing Date 

US10936177 Expired  Fee Related US7702002B2 (en)  20040128  20040907  Rapid acquisition methods and apparatus for GPS signals 
US12763026 Active US8111736B2 (en)  20040128  20100419  Rapid acquisition methods and apparatus for GPS signals 
US13353099 Active US8576895B2 (en)  20040128  20120118  Rapid acquisition methods and apparatus for GPS signals 
Family Applications After (2)
Application Number  Title  Priority Date  Filing Date 

US12763026 Active US8111736B2 (en)  20040128  20100419  Rapid acquisition methods and apparatus for GPS signals 
US13353099 Active US8576895B2 (en)  20040128  20120118  Rapid acquisition methods and apparatus for GPS signals 
Country Status (7)
Country  Link 

US (3)  US7702002B2 (en) 
EP (1)  EP1712010B1 (en) 
JP (1)  JP4616282B2 (en) 
CA (1)  CA2554412A1 (en) 
DE (1)  DE602005010391D1 (en) 
RU (1)  RU2360263C2 (en) 
WO (1)  WO2005074153A1 (en) 
Cited By (7)
Publication number  Priority date  Publication date  Assignee  Title 

US20050260947A1 (en) *  20040518  20051124  Karabinis Peter D  Satellite communications systems and methods using radiotelephone locationbased beamforming 
US20060114984A1 (en) *  20041117  20060601  Peter Gaal  Method and apparatus for increasing coherent integration length while receiving a positioning signal 
US20090110119A1 (en) *  20071025  20090430  Hitoshi Ikeda  Digital satellite broadcasting receiving tuner and satellite broadcasting receiving device including the same 
US20100135363A1 (en) *  20050301  20100603  Zhengdi Qin  Supporting a Signal Acquisition 
US8576895B2 (en)  20040128  20131105  Qualcomm Incorporated  Rapid acquisition methods and apparatus for GPS signals 
US8644788B2 (en)  20120502  20140204  Skytraq Technology, Inc.  Signal receiver, portable computer and signal acquisition method used in satellite communication system 
US20140072017A1 (en) *  20120911  20140313  Seiko Epson Corporation  Decoding method and receiving device 
Families Citing this family (40)
Publication number  Priority date  Publication date  Assignee  Title 

US7642959B2 (en) *  20030321  20100105  Global Locate, Inc.  Method and apparatus for processing satellite signals at a satellite positioning system receiver 
WO2004038944A1 (en) *  20021024  20040506  Nokia Corporation  Determination of the correlation phase between a signal and a replica sequence 
US7046193B2 (en) *  20031124  20060516  Honeywell International Inc.  Software GPS based integrated navigation 
US7471717B2 (en) *  20040517  20081230  Yi Ping Huang  Apparatus and method for acquiring spreadspectrum signals 
US20060239391A1 (en) *  20050420  20061026  Flanagan Michael J  Evaluating base station timing in an asynchronous network 
CN101243713B (en) *  20050811  20130925  艾利森电话股份有限公司  Method and arrangements in a mobile telecommunication network 
US7529236B2 (en) *  20050815  20090505  Technocom Corporation  Embedded wireless location validation benchmarking systems and methods 
US7605752B2 (en) *  20051228  20091020  Gabriel Technologies Corp.  Apparatus and method for controlling autonomous and assisted GPS modes in a wireless mobile terminal 
US8279910B2 (en)  20060302  20121002  Qualcomm Incorporated  Method and apparatus for code space search in a receiver 
EP1992080A1 (en) *  20060302  20081119  QUALCOMM Incorporated  Method and apparatus for code space search in a receiver 
US7847730B2 (en) *  20060927  20101207  Bae Systems Information And Electronic Systems Integration, Inc.  Software defined navigation signal generator 
US20080117104A1 (en) *  20061120  20080522  TsuiTsai Lin  Satellite tracking method 
US7456785B2 (en) *  20061129  20081125  Transcore Link Logistics Corporation  Navigation signal differential acquisition method and systems therefor 
US9088348B2 (en) *  20070618  20150721  Digi International Inc.  System and method for obtaining frequency and time synchronization in a wideband communication system 
US9088349B2 (en) *  20070618  20150721  Digi International Inc.  System and method for transmitting and detecting spread spectrum signals 
US7899126B2 (en) *  20070727  20110301  Mediatek Inc.  Signal receiver and method for estimating residual doppler frequencies of signals thereof 
US8179311B2 (en) *  20070821  20120515  Telecom Italia S.P.A.  Method for the acquisition of signals of a global navigation satellite system 
WO2009104278A1 (en) *  20080221  20090827  Fujitsu Limited  Filter device 
KR101446559B1 (en) *  20080324  20141006  삼성전자주식회사  Method to generate signal for watching 3D image and image watching apparatus thereof 
US7932859B2 (en) *  20080520  20110426  Motorola Mobility, Inc.  Method and apparatus for satellite positioning system time resolution 
US20110094445A1 (en) *  20080606  20110428  Ulvac, Inc.  Apparatus for manufacturing thinfilm solar cell 
US20100171657A1 (en) *  20080612  20100708  Northern Illinois University  System and method for wireless positioning and location determination 
CN101621739A (en) *  20080630  20100106  深圳富泰宏精密工业有限公司;奇美通讯股份有限公司  Signal acquisition system and method therefor 
US20100019967A1 (en) *  20080725  20100128  Farrokh Farrokhi  Method and apparatus for determining location 
US9857475B2 (en)  20080909  20180102  Geooptics, Inc.  Cellular interferometer for continuous earth remote observation (CICERO) 
US8625572B2 (en) *  20081219  20140107  Nokia Corporation  Synchronization indication in networks 
JP5500619B2 (en) *  20090330  20140521  日本電気株式会社  Gps terminal, a positioning method, communication system and program 
US7978130B1 (en) *  20090501  20110712  Coherent Navigation, Inc.  Practical method for upgrading existing GNSS user equipment with tightly integrated NavCom capability 
CN101937073B (en) *  20090630  20121128  凹凸电子（武汉）有限公司  Method and device for detecting carrier frequency of GPS (Global Position System) satellites as well as GPS receiver 
US8675712B1 (en)  20100408  20140318  Marvell International Ltd.  Zero delay reduced memory GPS PRN code interpolation 
US8958754B2 (en) *  20100929  20150217  Andrew, Llc  System and method for subcoherent integration for geolocation using weak or intermittent signals 
US8532112B2 (en) *  20110923  20130910  Lsi Corporation  Interleaving for wideband code division multiple access 
US8744518B2 (en) *  20120409  20140603  Apple Inc.  Mechanism for reducing time away from packet switched operation in a single radio solution 
JP5936902B2 (en) *  20120413  20160622  株式会社東芝  Transmission system, transmitter and receiver 
US9083444B2 (en)  20130312  20150714  Digi International Inc.  Chirp spread spectrum system and method 
US9766321B2 (en)  20130314  20170919  Digi International Inc.  Indoor positioning with radio frequency chirp signal propagation delay measurement 
CN103592663B (en) *  20131126  20151230  东南大学  A method of capturing GPS L2C code signal cl 
CN103941269A (en) *  20140415  20140723  南京邮电大学  PN code capturing method used for satellite navigation system 
CN105717522A (en) *  20160223  20160629  电子科技大学  Secondgeneration BeiDou B1 frequency band weak signal capturing method 
CN106556848A (en) *  20161024  20170405  华中光电技术研究所（中国船舶重工集团公司第七七研究所）  Rapid capturing method for signal of Beidou secondgeneration satellite B1 frequency point 
Citations (16)
Publication number  Priority date  Publication date  Assignee  Title 

US4093988A (en) *  19761108  19780606  General Electric Company  High speed frequency response measurement 
US5640431A (en)  19950310  19970617  Motorola, Inc.  Method and apparatus for offset frequency estimation for a coherent receiver 
US5781152A (en)  19951215  19980714  Sextant Avionique  Method and circuit for the reception of signals for positioning by satellite with elimination of multiplepath errors 
WO1999066649A1 (en)  19980615  19991223  Qualcomm Incorporated  System and method for narrowing the range of frequency uncertainty of a doppler shifted signal 
RU2144210C1 (en)  19980406  20000110  Научноисследовательский институт "Научный центр"  Sixchannel parallel correlation unit for receivers of satellite navigation equipment 
US6018312A (en)  19950920  20000125  The Secretary Of State For Defence In Her Britannic Majesty's Government Of The United Kingdom Of Great Britain And Northern Ireland  Locating the source of an unknown signal 
JP2000222389A (en)  19990203  20000811  Matsushita Electric Ind Co Ltd  Fourier transforming device 
JP2001237744A (en)  19991209  20010831  Nokia Mobile Phones Ltd  Receiver synchronizing method, receiver and electronic device 
WO2001089105A1 (en)  20000516  20011122  Koninklijke Philips Electronics N.V.  A method of despreading a spread spectrum signal 
WO2002013476A1 (en)  20000809  20020214  Skybitz, Inc.  System and method for fast code phase and carrier frequency acquisition in gps receiver 
US6407699B1 (en) *  20000414  20020618  Chun Yang  Method and device for rapidly extracting time and frequency parameters from high dynamic direct sequence spread spectrum radio signals under interference 
US6496533B2 (en)  19970925  20021217  Trimble Navigation Limited  Receiver having a memory search for fast acquisition of a spread spectrum signal 
US6735243B1 (en) *  20001108  20040511  Nokia Corporation  Method, apparatus and system for fast acquisition of a spread spectrum signal 
US6757610B1 (en)  20020204  20040629  Sirf Technology, Inc.  Method for rapid location determination using parametric values obtained in a previous location determination 
US20060023776A1 (en) *  20031211  20060202  ChiaChang Hsu  Correlator for spread spectrum receiver 
US7092433B2 (en) *  20010504  20060815  Asulab Sa  Radiofrequency signal receiver with means for improving the reception dynamic of said signals 
Family Cites Families (4)
Publication number  Priority date  Publication date  Assignee  Title 

DE4412876C2 (en) *  19940414  19960314  Borsi Kg F  A method of coating a regionwise, preferably transparent support plate 
US5852630A (en)  19970717  19981222  Globespan Semiconductor, Inc.  Method and apparatus for a RADSL transceiver warm start activation procedure with precoding 
JP4129088B2 (en)  19981015  20080730  株式会社日立ハイテクサイエンスシステムズ  Scanning transmission electron microscopy 
US7702002B2 (en) *  20040128  20100420  Qualcomm Incorporated  Rapid acquisition methods and apparatus for GPS signals 
Patent Citations (17)
Publication number  Priority date  Publication date  Assignee  Title 

US4093988A (en) *  19761108  19780606  General Electric Company  High speed frequency response measurement 
US5640431A (en)  19950310  19970617  Motorola, Inc.  Method and apparatus for offset frequency estimation for a coherent receiver 
US6018312A (en)  19950920  20000125  The Secretary Of State For Defence In Her Britannic Majesty's Government Of The United Kingdom Of Great Britain And Northern Ireland  Locating the source of an unknown signal 
US5781152A (en)  19951215  19980714  Sextant Avionique  Method and circuit for the reception of signals for positioning by satellite with elimination of multiplepath errors 
US6496533B2 (en)  19970925  20021217  Trimble Navigation Limited  Receiver having a memory search for fast acquisition of a spread spectrum signal 
RU2144210C1 (en)  19980406  20000110  Научноисследовательский институт "Научный центр"  Sixchannel parallel correlation unit for receivers of satellite navigation equipment 
WO1999066649A1 (en)  19980615  19991223  Qualcomm Incorporated  System and method for narrowing the range of frequency uncertainty of a doppler shifted signal 
JP2000222389A (en)  19990203  20000811  Matsushita Electric Ind Co Ltd  Fourier transforming device 
JP2001237744A (en)  19991209  20010831  Nokia Mobile Phones Ltd  Receiver synchronizing method, receiver and electronic device 
US7257153B2 (en) *  19991209  20070814  Nokia Corporation  Method of synchronizing a receiver, and a receiver 
US6407699B1 (en) *  20000414  20020618  Chun Yang  Method and device for rapidly extracting time and frequency parameters from high dynamic direct sequence spread spectrum radio signals under interference 
WO2001089105A1 (en)  20000516  20011122  Koninklijke Philips Electronics N.V.  A method of despreading a spread spectrum signal 
WO2002013476A1 (en)  20000809  20020214  Skybitz, Inc.  System and method for fast code phase and carrier frequency acquisition in gps receiver 
US6735243B1 (en) *  20001108  20040511  Nokia Corporation  Method, apparatus and system for fast acquisition of a spread spectrum signal 
US7092433B2 (en) *  20010504  20060815  Asulab Sa  Radiofrequency signal receiver with means for improving the reception dynamic of said signals 
US6757610B1 (en)  20020204  20040629  Sirf Technology, Inc.  Method for rapid location determination using parametric values obtained in a previous location determination 
US20060023776A1 (en) *  20031211  20060202  ChiaChang Hsu  Correlator for spread spectrum receiver 
NonPatent Citations (8)
Title 

International Preliminary Report on PatentabilityPCT/US2005/003540, International Search AuthorityIPEA/USAlexandria, VirginiaApr. 28, 2006. 
International Preliminary Report on Patentability—PCT/US2005/003540, International Search Authority—IPEA/US—Alexandria, Virginia—Apr. 28, 2006. 
International Search ReportPCT/US2005/003540, International Search AuthorityISA/USAlexandria, VirginiaApr. 15, 2005. 
International Search Report—PCT/US2005/003540, International Search Authority—ISA/US—Alexandria, Virginia—Apr. 15, 2005. 
Supplementary European Search ReportEP05712838, Search AuthorityMunichSep. 24, 2007. 
Supplementary European Search Report—EP05712838, Search Authority—Munich—Sep. 24, 2007. 
Written OpinionPCT/US2005/003540, International Search AuthorityISA/USAlexandria, VirginiaApr. 15, 2005. 
Written Opinion—PCT/US2005/003540, International Search Authority—ISA/US—Alexandria, Virginia—Apr. 15, 2005. 
Cited By (12)
Publication number  Priority date  Publication date  Assignee  Title 

US8576895B2 (en)  20040128  20131105  Qualcomm Incorporated  Rapid acquisition methods and apparatus for GPS signals 
US20050260947A1 (en) *  20040518  20051124  Karabinis Peter D  Satellite communications systems and methods using radiotelephone locationbased beamforming 
US8265549B2 (en) *  20040518  20120911  Atc Technologies, Llc  Satellite communications systems and methods using radiotelephone 
US20090137203A1 (en) *  20040518  20090528  Atc Technologies, Llc  Satellite communications systems and methods using radiotelephone locationbased beamforming 
US8254512B2 (en) *  20041117  20120828  Qualcomm Incorporated  Method and apparatus for increasing coherent integration length while receiving a positioning signal 
US20060114984A1 (en) *  20041117  20060601  Peter Gaal  Method and apparatus for increasing coherent integration length while receiving a positioning signal 
US20100135363A1 (en) *  20050301  20100603  Zhengdi Qin  Supporting a Signal Acquisition 
US8254499B2 (en) *  20071025  20120828  Sharp Kabushiki Kaisha  Digital satellite broadcasting receiving tuner and satellite broadcasting receiving device including the same 
US20090110119A1 (en) *  20071025  20090430  Hitoshi Ikeda  Digital satellite broadcasting receiving tuner and satellite broadcasting receiving device including the same 
US8644788B2 (en)  20120502  20140204  Skytraq Technology, Inc.  Signal receiver, portable computer and signal acquisition method used in satellite communication system 
US20140072017A1 (en) *  20120911  20140313  Seiko Epson Corporation  Decoding method and receiving device 
US8885687B2 (en) *  20120911  20141111  Seiko Epson Corporation  Decoding method and receiving device 
Also Published As
Publication number  Publication date  Type 

WO2005074153A1 (en)  20050811  application 
JP2007519936A (en)  20070719  application 
RU2006130799A (en)  20080320  application 
US20130022085A1 (en)  20130124  application 
DE602005010391D1 (en)  20081127  grant 
US20050163201A1 (en)  20050728  application 
JP4616282B2 (en)  20110119  grant 
CA2554412A1 (en)  20050811  application 
US8576895B2 (en)  20131105  grant 
EP1712010B1 (en)  20081015  grant 
US20110103433A1 (en)  20110505  application 
US8111736B2 (en)  20120207  grant 
RU2360263C2 (en)  20090627  grant 
EP1712010A1 (en)  20061018  application 
EP1712010A4 (en)  20071031  application 
Similar Documents
Publication  Publication Date  Title 

Braasch et al.  GPS receiver architectures and measurements  
US6816710B2 (en)  Method and apparatus for signal processing in a satellite positioning system  
US6295023B1 (en)  Methods, mobile stations and systems for acquiring global positioning system timing information  
US6131067A (en)  Clientserver based remote locator device  
US6208291B1 (en)  Highly parallel GPS correlator system and method  
US6211820B1 (en)  Call maintainance during position location  
US6952440B1 (en)  Signal detector employing a Doppler phase correction system  
US6583757B2 (en)  Method and apparatus for satellite positioning system (SPS) time measurement  
US6346911B1 (en)  Method and apparatus for determining time in a GPS receiver  
US7545894B2 (en)  Method and apparatus for detecting and processing global positioning system (GPS) signals  
US7158883B2 (en)  Method and apparatus for locating position of a GPS device  
US7764726B2 (en)  Systems, methods, and apparatus for jammer rejection  
US20040071200A1 (en)  System for direct acquisition of received signals  
US7042930B2 (en)  Spread spectrum bit boundary correlation search acquisition system  
US6829534B2 (en)  Method and apparatus for performing timing synchronization  
US6735243B1 (en)  Method, apparatus and system for fast acquisition of a spread spectrum signal  
US20090251364A1 (en)  Method and system of a mobile subscriber estimating position  
US20030161543A1 (en)  Spectrum spreading signal demodulation method and apparatus  
US6633255B2 (en)  Method for open loop tracking GPS signals  
US6285316B1 (en)  Locating a mobile unit using signals from both mobile beacons and stationary beacons  
US5812087A (en)  Method and apparatus for satellite positioning system based time measurement  
US6909738B2 (en)  Method in a receiver and a receiver  
US20050265494A1 (en)  Method and apparatus for code phase tracking  
US20070008217A1 (en)  Method and device for acquiring weak global navigation satellite system (GNSS) signals  
US20060055596A1 (en)  Satellitebased positioning system improvement 
Legal Events
Date  Code  Title  Description 

AS  Assignment 
Owner name: QUALCOMM INCORPORATED, CALIFORNIA Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:KRASNER, NORMAN F.;REEL/FRAME:015484/0175 Effective date: 20041207 Owner name: QUALCOMM INCORPORATED,CALIFORNIA Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:KRASNER, NORMAN F.;REEL/FRAME:015484/0175 Effective date: 20041207 

CC  Certificate of correction  
FPAY  Fee payment 
Year of fee payment: 4 

FEPP 
Free format text: MAINTENANCE FEE REMINDER MAILED (ORIGINAL EVENT CODE: REM.) 

LAPS  Lapse for failure to pay maintenance fees 
Free format text: PATENT EXPIRED FOR FAILURE TO PAY MAINTENANCE FEES (ORIGINAL EVENT CODE: EXP.) 

FP  Expired due to failure to pay maintenance fee 
Effective date: 20180420 